BF Fender VibroVerb and Deluxe Reverb double note intermod at 15 Fret

[Joe L]

I've built clones of the two amps in the subject line. Both are true to the BF schematics and I have the equipment to very the parameters on all the components and do so before they are installed. They sound great except for the fact that both produce intermod double noting around the 15 fret - second string = E when driven into power amp distortion with nothing but a Les Paul with a SD Pearly Gates on the rear. My mid '70s 100 watt Marshalls do the same thing with new PS caps. I know it is not cone cry because it is independent of speaker type/cabs. I can and have tried adding first stage filtering (from 35uf to 105uf on the VVerb) to reduce the intermod but I really don't want to stiffen the supplies. Any suggestions?

[bob p]

I have to be really brief because it's about supper time here. Did you search the forum for "ghost note"?

http://music-electronics-forum.com/s...searchid=14989

[loudthud]

Have you tried a different guitar?

[J M Fahey]

I bet that note is at or real close to a 60 Hz harmonic, i.e. 120 or 180 Hz.
It's normal.

[Joe L]

I have tried other guitars, and like I say, it happens also on my '70s 100 watt Marshalls. I just did a search for ghost notes and like I suspected and what was suggested, its intermod with the supply ripple.

Even though it is normal, I am going to do some experiments to lower the ripple while keeping the impedance of the supply about the same.

I'll let you guys know what if anything I find. In the meantime, comments are welcome.

[Joe L]

Just an update.

I found a thread from '07 where Chris Merren brought up the fact that one cause of intermod/double notes/hum at full tilt can be the dynamic unbalancing of the PI output when pushed into clipping. This contributes to the asymmetrical bias shift caused by grid current of the finals charging up the coupling caps from the PI. The bias shift is present in almost any amp when driven hard into distortion of the PA but as long as it is balanced, the stage doesn't get into full blocking distortion. Chris says that it is easy to see if this is happening by measuring DC voltage between the junction of the bias resistors and PI coupling cap. One lead on each side, not to ground reference. Theoretically you should read 0vDC all the way to hard clipping in a perfect amp which of course, doesn't exist.

The point of this is if everything is reasonably balanced, common mode signals to the PA should be cancelled. Loose the balance for any reason, hum and intermod skyrocket.

With this in mind, this is what I have thought up so far in an attempt to fix the problem.

1. Use a balanced PI tube
2. Raise one or both of the grid resistors on the PA to increase and better balance the time constant of the PI caps/grid resistor network.
3. Lower the value of the bias resistors to shorten the time constant with the PI coupling caps.
4. Adjust the plate resistors to better balance the PI output.
5. Raise the value of the PI tail resistor to better balance the circuit output.
6. Change the feedback resistor value. More feedback may help or hurt, I'm not sure about this but it is easy to put a temporary pot.
7. Raise the value of the screen resistor on the "hot" side

I have another project on the bench that I can't drop at the moment to experiment with this, but I just wanted to maybe get some discussion going. I know I'm not the only one that has experienced either this, or a Marshall that redplates outputs only on one side when driven hard.

[Steve Conner]

This kind of bias shift phenomenon is typical of the things that tube amps do and solid-state ones don't. The asymmetrical clipping and bias pumping adds interest to the overdriven tone. If you fixed all the sources of bias shift, you might as well use a solid-state amp.

One possibility would be to use a large choke to filter the entire B+. Adding a choke gets more ripple reduction without increasing the filter capacitance. You just split it in two, half before the choke and half after.

[Joe L]

This kind of bias shift phenomenon is typical of the things that tube amps do and solid-state ones don't. The asymmetrical clipping and bias pumping adds interest to the overdriven tone. If you fixed all the sources of bias shift, you might as well use a solid-state amp.

One possibility would be to use a large choke to filter the entire B+. Adding a choke gets more ripple reduction without increasing the filter capacitance. You just split it in two, half before the choke and half after.

I agree about fixing all the sources. The sound of the PA would be like a diode clipper. But obviously there are Marshalls that don't redplate one side and sound fantastic so there is some wiggle room for tweaking without killing the tone. And a large part of this is just a mental exercise to learn more about output stages.

First thing this morning I did a bit of research and found calculators over at Ampbooks.com for LTP and Bias Excursion tweaking. I also visited Aiken's page on LTPs and did some more reading. From the info I've got so far, it looks like tweaking the plate resistors of the PI with a pot is the best place to start. The amp I am building for a friend that is on the bench (a dual identical clean/crunch channel SLO) is just at the stage where I have all the preamp tubes in and hot. I can use it to tweak a bit while doing my scope inspection of the build.

And the large choke just might be a good solution but a lot more expensive that a few resistors . But I do have a choke input screen filter on my list of things to play with in the future.

[loudthud]

'umble amps have a trim pot to balance the load resistors of the PI. There is much speculation about how and why the adjustment is made. Let us know how it affects you problem.

For a choke, you don't need a very big one. Something less than 1 Henry could get the job done but look at the current rating and keep the resistance below 50 ohms.